Although a dose-response curve clearly exists for alkylating agents in the initial chemotherapy of small cell lung cancer, the therapeutic benefit of higher than standard doses of the more recently introduced regimen of etoposide/cisplatin (VP16/PLAT) is uncertain. We randomized at least partially ambulatory patients with extensive stage SCLC and without major organ dysfunction to receive either VP16 80 mg/m squared + PLAT 27 mg/m squared Days 1-5 q 3 wks or VP16 80 mg/m squared Days 1-3 + PLAT 80 mg/m squared Day 1 q 3 wks for the first 6 wks of therapy. Nonambulatory patients and those with organ dysfunction were assigned standard dose treatment. All patients received the standard dose regimen during wks 7- 12. From wks 13-24, patients in complete response (CR) continued standard dose VP16/PLAT, while all other patients received a new 3-drug regimen that led to further improvement in response in only 5 cases. CR's were given prophylactic cranial irradiation. Ninety-eight patients have been entered (77 of whom were randomized). CR rates (p=1.00) and survival (p=0.73) were similar in patients randomized to high and standard dose therapy. There were 2 treatment-related deaths in the high and one in the standard dose arm. We conclude 1) standard dose VP/PLAT is at least as active as any regimen we have ever utilized for extensive stage SCLC and produces only modest myelotoxicity, and 2) there is no evidence of superior efficacy when planned drug doses are increased by 67% during the first 6 wks.